O'Reilly Tools of Change for Publishing, in San Jose. Here (so far) are raw notes from the tuesday sessions:

Here's the conference program: http://conferences.oreillynet.com/toc2007/schedule/

See more notes at TOC07 Day 2 Notes

Here we are at TOC. So far it seems like there's only O'Reilly employees here, but I suspect that's just sampling error on my part. Maybe 300+ people here. 200 people typing in a big room sounds remarkably like rain on the roof. Neat!

Sarah Milstein, program chair, begins, with Tim O'Reilly doing the history of publishing tech, beginning with the Sumerian clay tablet and the origina DRM (the book curse). Woodblocks 868BC - the Diamond Sutra. Roman codex: Codex Abrogans - 1AD. And the medieval helpdesk video. Gutenberg's moveable type. Ben Frankin - "Pioneer or Pirate"? Dickens quote on the importation of same-langauge books to America ("a greater curese cannot be inflicted"). Bush's Memex 1945. Macintosh, Pagemaker, Laserwriter 1985. Encyclopedia on CDROM 1985. Berners-Lee 1991. Phil Greenspun (Phillip and Alex's Guide to Web Publishing) 1998. Wikipedia 2001.

Brian Murray - "Retooling HarperCollings? for the Future." Since 2004, a major shift in HC's operations. "Book search" before 2004 was called a "bookstore". Alexa numbers show that since 2006, Google's traffic is growing, while Amazon's is falling off. 2002 NEA numbers show power distribution between number of books read by people. HC interested directly in the 90million American who read more than 6 pr year, but Google, Yahoo also catch the rest, who read fewer. Half of HC's web traffic is first-timers coming from search -- self-described book enthusiasts. 26% were searching for authors. Only 6% of incoming traffic goes to homepage.

Direct sales not a priority just now, but digitizing books and building "digital warehouse" are. 2005 was "tipping point year"—realized that they didn't want multiple copies floating around, but to be the source of the "exclusive digital copy". There are now "print consumers" and "digital consumers". HC's 2006/07 digitization a major publicity move for them. LibreDigital? selected as partner in creating the "warehouse".

Text processing, workflow management, digital asset management add up to savings of >$4000 per title to boot. 12000 titles now in warehouse, 6000 on website, representing the high end and the low end of their lines first. Browse Inside widget and app has been a success with bloggers and myspace; Oprah site picked up the widget; newsletters too. Overall strategy is infrastructure -> new marketing -> driving traffic -> new products.

Chris Anderson on his new book Free. What Moore's Law really means: the price of a resource ultimately falls to zero. Changes from something you conserve to something you (should) waste. So Skype, GMail; Yahoo now offering infinite storage. Economy increasingly running on significant non-monetary bases. If the marginal cost approaches zero, then treat it as zero and get ahead.

Broadcast technology works this way: marginal cost of additional consumers is zero, which changed culture hugely: rise of mass markets; we synchronized our culter around the economics of broadcast, around amortizing capital costs, hence blockbuster and homogenization. Magazines and newspapers are essentially the same, because of CPM model. Subscriptions are simply a device to qualify readers to advertisers. Cross-subsidized product, making the money back in advertising.

Most music is already free now, both monetarily and rights-wise. Artists want to give it away to promote themselves, build audiences for performances, etc. Books are the last media not seeming to be approaching free. Why? Books "make sense". Physical product is better than the digital product.

So in what ways can a book be free? Audiobook version can be free, especially (only?) for people who buy the book. Booksearch is free. E-book locked to specific reader can be free. Unlocked ebook with ads can be free, similarly web-browsed version. Free samples are free. Marginal costs of all are all nearly zero, so strategically we treat them as zero.

Maximizing audience/reach is more important for the vast majority of authors than maximizing revenue, given 200000 new books a year.

How to make the physical book free? Sponsored books are not new, nor even rare. Consultants give away books as promotion all the time. Ads in books is not new nor rare (esp. pulp); this is all about proper targeting, of course. Free rebates in exchange for personal information/personal relationship. Free promo copies—free copies of The Long Tail to bloggers and reviewers drove so much traffic to Amazon that online sales outpaced Barnes and Noble sales. And of course libraries.

MAKE magazine has ad-embedded PDF downloaded; 80000+ downloads, 5% clickthroughs!

Free media, act as promotion for the person or presence, which is scarce and therefore expensive, or promotion for the superior thing, like a printed book. Free maximizes reach to the "new influentials"

Why aren't there more free books? Because few authors can monetize their fame (as CA can). Excerpting doesn't always work. Advertising is difficult to match to long works. Publishers are still in the manufacturing and distribution business, which is "misaligned" with Chris Anderson's own celebrity marketing. Books not yet seen as a part of a larger community, a larger discussion. Channel conflict worries are a serious disincentive to innovation.

Tim O'Reilly on Web 2.0. O'Reilly's business is "changing the world by spreading knowledge of innovators". Future here, not evenly distributed yet (Wm Gibson). "Harnessing collective intelligence" a better term than "user-generated content". Systems that harness network effects. Network effect-driven data lock-in, because of scale. Free is a precursor of rediscovery of value in other forms: a huge challenge to publishers to find where the value is going. Music companies missed this and gave a lot of it to Apple.

Curating user-generated content is the "new face of publishing". Allowing users to re-organize content. (This is pretty similar to what Kim Pittaway was talking about at MagNet?, about repurposing content for the web). O'Reilly Cookbooks are modular, structured, and invite recipes from users, which are then edited and published.

What job do books do? Different things. Reference books do what other search things do; access and organization of information. Safari program is 3rd only to Amazon and B&N for O'Reilly revenue. Search changes the reading experience. People read less than 5% of any book in a month. Long tail drives access to older content. There is lots of Safari traffic way, way out the tail... things that are just not selling at all any more. Google book search pattern is the same... people are digging for stuff that isn't sellingin print.

"Piracy is progressive taxation"—it takes a bit from the bestselling works, but it really brings the bottom up. The question is, how do you manage this to your advantage?

Chris Anderson and Tim O'Reilly talk a little about behind-the-curtain content online. CA says for WIRED, the conversion rate of free access to that makes way more sense than holding it for subscribers only. TO'R points out that for smaller scale publications, the trade-offs will be different, and so it is critical to "instrument" your site to see where things are happening and be able to make changes quickly. CA: "fail fast... and cheaply".

TO'R points out that Harry Potter and World of Warcraft are competing properties.

Session on DRM:

Peter Brantley from Digital Library Federation on what is DRM for, really? Minimizing risk, not preventing it. Creating stickiness. Content we have not just increasingly fragmented, but also increasingly interconnected with new sources. What do a access restrictions do in that kind of environment, as opposed to the traditional "large objects" approach?

Ale de Vries from ScienceDirect?/Elsevier. ScienceDirect? has 4 million articles over the past decade, plus another 4 million of old, digitized content. Also a book series: 150 titles, plus 53 reference works, plus series handbooks, and 4000+ ebooks (mostly textbook-like things). Search, personalization, the usual portal things. Big, blanket licenses, with the goal of having seamless access to users. Trusting librarians is key. Need to do monitoring to watch for abuses.

Michael Jensen from National Acadamies Press, a press, a research service, a policy recommendation system. Originally sold content per page; still roughly based on this model. Free browsing of pages online, married with sales of complete books. Balance between openness and self-sustainability is key challenge. Continual work on improving online UI to allow people to find things. Openness expands the audience enormously. Only 0.2% conversion to book sales still made $2 million, or 1/3 of their business. Openness is working as a business decision. Reading online, even PDFs?, is non-optimal enough that it operates more as publicity than cannibalization of sales. "Erring on the side of protection will kill you faster than erring on the side of openness." Don't underestimate the "deep niche" revealed by patterns of usage of low-profile titles over time.

Peter Brantley asks the other two about the self-containing limits of non-restricted content. Ale de Vries tells of one guy who attempted to commercially pirate ScienceDirect? stuff, but of course they immediately knew about him, and were able to shut him down. It's an annoyance, but it doesn't really hurt the bottom line. Michael Jensen says the odd professor who puts NAP's content into Blackboard is outweighed by the expanded audience that viral dissemination brings.

Peter Brantley says, the cardinal rule of business is to keep your customer satisfied. Why would you want to get in the way of that? Ale de Vries says that the complications that DRM brings to customer service and business development are considerable. What are you trying to do, exactly?

On a question, it turns out that no one in the room of 40 people is doing DRM on PDFs?.

Peter Brantley responds to a question about the kind of tracking and re-flowing that performers' rights organizations might make sense for large aggregators to take on, and predicts that we will start to see this.

Asheesh Birla, from Thomson Learning on using wikis. Context is downward trend in textbook prices, rise of Wikipedia and Google, custom publishing on the rise, need to do much faster production cycles. Wikis as a possible way to go do faster/cheaper?

Wikis as internal tool: project planning, content creation, distribution. Reference division. Twiki + project dashboard for project planning. Wiki use brought out te "long tail" of junior employees participating in producing plans, not just the people that a word doc gets circulated to.

Used Active Collab (like BaseCamp?) to organize content creation—reference content. Moved away from editor being the hub for multiple content authors and sources and to a tool in the middle (this is the same as the OJS model); the editor becomes a manager or co-ordinator. Tags used as early start of formal metadata creation process. Smaller project components form the locus for workgroups, rather than an entire project. Author training was required, and ongoing support from editors.

Question comes up about change-tracking. Asheesh: this takes some getting used to, and needs more attention in future development.

Content distribution, new business... Wikipedia moves so much faster than reference publishing. Market now expects this responsiveness.

Getting world online rights to images and media is very expensive. Wikis and Web2.0 tools like Flickr provide at least a path to sourcing content.

Bringing authors into a longer-term content maintenance context is great; needs to be addressed in contracts, of course. Custom publishing is a nice (quick) application for wikis in content management.

Lots of questions about education, eLearning, corporate buy-in.

Session on publishers' websites:

Maja Thomas from Hachette Digital Media reports that they de-emphasized publisher and imprint, and led with authors instead. Discusses the usually hookysticky tricks and gadgets that modern web marketers employ: contests, call outs, widgets, podcasts, games, rotating see-alsos (it occurs to me a lot of this is scattershot design: throw every bell and whistle on and see what draws attention).

Brian Napack from Holtzbrinck Publishers, a publisher with very high-profile brands, and lots of higher-ed stuff. "Publishers help creators find and connect with audiences through their work." And also, "We don't need publisher websites." The real traffic comes qualified, through to product sites (authors, series, other brands). Book distribution is profoundly broken; we need to get beyond the bookstores, because most people don't actually ever go into them. Need asset-centric websites, not publisher-centric websites. Holtzbrinck didn't decide to publish Mignon Fogarty's Grammar Girl (which was a podcast); rather, they bought Fogarty's media brand. the book isn't even out yet. Holtzbrinck also hosting book-centric social networking site lovelybooks.de

(Napack's message seems to be that it's not about books and websites as discrete things; it's about managing media brands across the board.)

Rikard Linde talks about "why people buy things they can get for free," and begins with the story of Evian, the water company. Bottled water is a $35 billion industry. Of all the reasons people buy water, one of the interesting reasons is that it guarantees purity in a unique way. Cites John Perry Barlow's (1994) "Economy of Ideas" and the argument about wine vs wine bottles. Talks about iTunes guaranteee of the right song at high quality, with no hassles; this is the proposition over filesharing. So... the proposition with books, as Chris Anderson alluded to this morning, is the whole book, the real book, the correct book, the genuine article. There is probably a bunch more value publishers could include that would make the paid-for article even more interesting.

The Gary Troup (from LOST) book, Bad Twin, has (actually) been released by Hyperion. Of course, it has all sorts of other virtual/real/virtual associations. The Hanso Foundation website actually denounces the book. So what is real, what is authentic here? But LOST fans are interested in it, and a sort of authenticity accrues. You can sell a sort of guarantee that a book relates to a topic, even though the relation might be pretty tenuous.

Jimmy Wales talks about Wikia an all-GPL based library of wikis. This is about the "Free culture" movement. "Second wave" of collaborative content: There are 300 articles on the Muppets on Wikipedia, but Muppet Wiki (part of Wikia) has 14,000 articles. Not NPOV, but a million (well, 3000 so far) different points of view. Wikia aspires to be a sustainable, user-controlled media company (different from Wikipedia foundation): advertising supported.

Search Wikia aims to create a transparent, free-software search engine. This should shift power away from search companies and back to content companies, he says.

Spontaneous random contribution/collaboration is more important than security. We eat in restaurants next to people weilding steak knives, and it's not a problem.

Jeff Patterson from Safari. "How much do we know about our subscribers?" Safari has upwards of 40000 subscribers, and measures monthly usage in the hundreds of thousands of 12-18 minute sessions.

Three currencies: money, time, personal information. Surveying shows that over 57% of surveyed are not happy paying with personal information. And the more detailed the personal info, people are even less comfortable. People would prefer to simply pay money.

Bruce Chizen of Adobe with Tim O'Reilly. TOR asks about convergence of print and web side. Launching Digital Editions today, and later this year, Adobe Integrated Runtime (AIR, formerly known as Apollo). Blending PDF, HTML, and Flash, without being inside the browser. The plugins break out. He positions DRM as a document versioning and lifecycle tool. However companies want to use DRM, we're providing it. TOR: "you're kind of an arms-dealer, then." BC: Adobe is a company built on IP; we take it seriously. TOR: you're also releasing things like Flex as open source. BC: it all depends on the business model, but I want that to be my choice.

TOR presses BC about platforms, and the iPhone, and authoring tools, and so on. He doesn't really say anything, except to reinforce the idea that Adobe is at the centre of web authoring and print authoring, and there are "opportunities."

Last word: "Experiment. Print will go away eventually."